internet energy consumption 2
This is a follow-up to my post of
29 October. After having watched the Newsnight programme I can only conclude that matters are still vague if not sketchy with regard to a "green Internet". The key terms (buzzwords / -phrases) are "cloud computing", "virtualisation", and "follow the moon". The first two you´ve probably heard about by now, since they have been bandied about for at least the past 18+ months.
Since I´ve always written my blog with a non-specialist audience in mind, that audience-notion now becomes muddled as I have the postings fed on over to the feed at the
OII. But let me explain anyway, in a way that I might have done had I still been teaching undergraduate ICT courses.
- Recall the early mainframes-and-dumb-terminals computing model. Now transfer that notion to the Web, where we now have Web services (think Gmail, GoogleDocs, etc). Cloud computing differs from Web services (conceptually) in so far as the number and diversity of services that can be delivered from "the cloud".
- Virtualisation can be understood as the maximisation of computing resources (at the level of the platform, processing resources, and/or application) (think here "one box, delivering many and disparate services or functions).
- The term "follow the moon" can be understood as a way in which to distribute energy-hungry computer processing to areas of the planet where energy-demand is at its relative lowest e.g. if demand for computer processing is high, say in daytime London, let the energy needs for that processing be met from a spot anywhere on the planet where it is night-time. (Two additional methods not mentioned in the tv programme, were "follow the sun", and "follow the law". See more on this in Kevin Kelly´s post.)
- Another option proferred in the programme was the location of server farms on (very) cold spots on the planet e.g. Iceland, which does away with the need to manufacture cooling of these farms.
So can we imagine then a world where all of the above solutions were applied? And if so, would that lead to significant reductions in CO2 emissions from the Internet industries and our collective computing activities? That still has to be demonstrated. For now, what is said is that already the CO2 emissions from our computing activities equal that of the car manufacturing industry, and are set to match those of the airline industry in 2020.
As an aside: Cloud computing is not uncontroversial. See, for instance, the piece "
Cloud computing is a trap, warns GNU founder Richard Stallman (The Guardian newspaper)"
See also:
Guide To Cloud Computing (InformationWeek) ;
Virtualization (Wikipedia).
Labels: energy consumption, internet, OII
Two years ago I blogged on the energy consumption needs of the Internet, making the point that if African countries truly wanted to compete, how were they to do so if they (more than others) lacked scarce energy and water resources. My thinking then had been prompted by an article on massive server farms /data stores in the August 28, 2006 issue of Fortune magazine (European edition). Of interest still are the questions I´d posed in my
1-Sept-2006 posting, which rang as follows:
Okay, so, the article "The future of computing (part one)". Some excerpts:
"Most people don´t think of it this way, but the Information Age is being built on an infrastructure as imposing as the factories and mills of yore....To handle this change [of software becoming webified], Internet companies are building their own [data] centers..."
And what data centers need are:
- ground, acres and acres of it
- electricity, not only to operate the servers but also to cool all those processors chugging away (e.g. "...for every dollar a company spends to power a typical server, it spends another dollar on a/c [air-conditioning] to keep it cool.")
- water, for cooling purposes, as increasingly alternative means are being explored to keep server farms cool.
moral of the story?
- even in the information age, we come back to the same basic ingredients for the infrastructure needed at base.
- seriously, where does this leave Africa in the race to be part of the Info Age?
greater moral of the story?
we are running low on fossil fuels, and water, on the planet (among other things, admittedly). the info age was supposed to signifiy a reduction in the demand for either of the two. instead, demand is only increasing. so, where does that leave us, after all?
It seems that my parting question of the energy-hungry Internet is to be addressed in this evening´s (Wed 29 Oct 2008) edition of
Newsnight (a BBC current affairs programme). Susan Watts, the Science correspondent of said programme
has blogged on the matter in anticipation of tonight´s episode. Reading the blog, it seems there is no clear answer as yet to how Internet companies expect to "go green", her having consulted IBM, Cisco, and Google. Thát, or maybe she´s saved the full answers for her broadcast... Once I´ve watched the episode on energy-hungry iplayer, I´ll let you know.
Labels: energy consumption, internet, OII
is social study of the Internet akin to writing a history of the present?
This weekend I´ve read two history books (Beevor on the spanish civil war, the other Garton Ash on Europe in the 1990s). And it is prompted by Timothy Garton Ash´s musings in the introduction to the volume named "History of the Present", that I now write here. I mention both books since they seem to deal with two different kinds of History scholarship: the one more traditional, the other not so; and so formed an interesting contrast for me.
In said introduction Garton Ash contemplates if not elucidates the differences between writing a history of the present versus writing journalism or say literature. What separates the first two from the third is "the kind of truth being sought" and that, for both History and Journalism the facts cannot be played "fast and loose" with. For one, he says that Historians always need at least thirty years to elapse before they tend to study an era or event what by then has been deemed to be "historical". And since his (Garton Ash´s) works have always been more about "the present", in the interstice between recording recent history as historian versus journalistic recording of what has happened, he goes on to distinguishing the two professions from one another, and in turn, from literature. As I read this intro it made me think about what differentiates good scholarly work in the area of studying the Internet, versus something as being more journalistic. For, after all, in studying technology, and in particular the Internet, are we not engaged in an exercise of writing, in a mode similar to Garton Ash, a kind of "history of the present"? In this regard, I wish Garton Ash would write a manual of praxis for his art/science, if only so that I could ponder more on what it can mean to do "social studies of the Internet". But then again, maybe my stance contradicts his closing "argument". He writes and concludes as follows:
"So I maintain that, for all its pitfalls, the literary enterprise of writing ´history of the present´ has always been worth attempting. It is even more so now, because of the way history is made and recorded in our time. And it has suffered from developments in the professions of journalism and academic history. Yet you can soon have enough of such methodological self-examination. Altogether, the habit of compulsive labelling, pigeon-holing and compartmentalizing seems to me a disease of modern intellectual life. Let the work speak for itself. In the end, only one thing matters: is the result true, important, interesting or moving? If it is, never mind the label. If it isn´t, then it´s not worth reading anyway."
(Garton Ash, T. 1999. History of the Present: Essays, sketches and despatches from Europe in the 1990s)
Links FYI publisher/purchasing info:
Garton Ash
Beevor
Labels: history, internet, nature-of-what-we-do
